| From: | Christophe Pettus <xof(at)thebuild(dot)com> |
|---|---|
| To: | Alanoly Andrews <alanolya(at)invera(dot)com> |
| Cc: | Ninad Shah <nshah(dot)postgres(at)gmail(dot)com>, "pgsql-general(at)lists(dot)postgresql(dot)org" <pgsql-general(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: Streaming replication versus Logical replication |
| Date: | 2021-11-04 20:55:20 |
| Message-ID: | 55F01B76-E5A6-47DD-8247-47A5835F2CBA@thebuild.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
> On Nov 4, 2021, at 12:16, Alanoly Andrews <alanolya(at)invera(dot)com> wrote:
>
> Thanks, Ninad, for the response.
> So, am I to understand that when there is a long-running query on the subscriber, the vacuumed data updates from the publisher are held over on the subscriber until the query completes? If so, where and how are they held over, and what does it mean in terms of disk space (either on the publisher or on the subscriber)?
That's not quite correct.
Vacuum operations happen independently on the publisher and subscriber in logical replication. Unlike binary replication, changes caused by vacuuming on the publisher are not sent over to the subscriber.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Adrian Klaver | 2021-11-04 22:30:27 | Re: to_date() and to_timestamp() with negative years |
| Previous Message | Alanoly Andrews | 2021-11-04 19:16:20 | RE: Streaming replication versus Logical replication |